You are on page 1of 8

International Journal of Mass Spectrometry 413 (2017) 168–175

Contents lists available at ScienceDirect

International Journal of Mass Spectrometry


journal homepage: www.elsevier.com/locate/ijms

Parallel accumulation for 100% duty cycle trapped ion mobility-mass


spectrometry
Joshua A. Silveira a , Mark E. Ridgeway a , Frank H. Laukien a , Matthias Mann b ,
Melvin A. Park a,∗
a
Bruker Daltonics Inc., Billerica, MA 01821, United States
b
Proteomics and Signal Transduction, Max-Planck-Institute of Biochemistry, Am Klopferspitz 18, 82152 Martinsried, Germany

a r t i c l e i n f o a b s t r a c t

Article history: A new trapped ion mobility spectrometry (TIMS) apparatus has been developed and evaluated. Building
Received 5 February 2016 on the original prototype design, the new spectrometer includes an upstream “ion accumulation trap”.
Accepted 7 March 2016 Inclusion of the new trap allows for storage of a population of ions during the analysis of a previously
Available online 14 March 2016
stored population of ions. By accumulating and analyzing ions in parallel, rather than in series, a duty
cycle of 100% can be achieved. The new design retains the flexibility of the original design—the ability to
Keywords:
exchange mobility range, resolution, and speed/duty cycle—while in all instances, maintaining superior
TIMS
ion utilization efficiency. When operating at 100% duty cycle, the spectrometer is shown to trap a broad
Ion mobility-mass spectrometry
Duty cycle
range of ions (m/z 622–2722) with an average trapping efficiency of ∼70%. Trapping efficiency was found
Parallel accumulation to be constant over a range of accumulation/analysis times from 20 up to 85 ms where an average resolv-
Charge capacity ing power of 60 was achieved. Compared to transmission mode (MS-only) operation, the combination
of high mobility resolving power, high ion utilization efficiency, and relatively fast scan times achieved
with TIMS is shown to result in an improvement in mass spectral peak intensities of more than an order
of magnitude.
© 2016 Elsevier B.V. All rights reserved.

1. Introduction end, while ion-neutral collisions give rise to an opposing steady-


state drag force. Separation results from a disparate number of
Relative to mass spectrometry (MS), the field of ion mobil- collisions experienced by ions of different mobilities (K). During
ity spectrometry (IMS) has been slow to mature. Early reports of drift, spatial broadening of the ion swarm stems from thermal dif-
hybrid IMS-MS instruments used to characterize the physical prop- fusion and Coulomb repulsion which principally limit the resolving
erties of gaseous ions began to emerge in the 1960s [1,2]. IMS was power and the number of ions that can be analyzed.
introduced in the 1970s as a rapid separations technique—“plasma Modern dispersive IMS systems are routinely coupled to time-
chromatography”—and has been widely used thereafter as a stan- of-flight (TOF) MS, owing to the relative speed of the two
dalone chemical sensor [3]. Still, only in the last couple of decades techniques. That is, IMS analysis typically occurs on a timescale
have a number of technological advances and commercial sys- between 10 and 100 ms. Assuming a resolving power of 50, IMS
tems extended the use of IMS-MS to a broad range of applications, peaks will have a temporal width between 0.2 and 2 ms. In contrast,
including detection of explosives and chemical warfare agents [4,5], TOF MS analysis only requires between 0.1 and 0.2 ms ensuring
characterization of polymers [6,7], cluster assemblies [8–12], and that 1–20 mass spectra can be acquired during elution of an IMS
biomolecular structures [13–25], as well as high throughput sepa- peak. Thus, hybridization of IMS with TOF MS provides a means for
ration of complex mixtures [26–32]. efficiently handling the ions to produce IMS-MS spectra.
Dispersive IMS is conventionally carried out in a drift tube that In the 1990s, a significant improvement in IMS-MS sensitivity
contains an inert buffer gas. During IMS analysis, a weak electric was achieved when radio-frequency (RF) ion funnels were inte-
field guides ions injected at one end of the tube toward the opposite grated into hybrid IMS-TOF MS instruments [33,34]. Developed in
response to then-emerging “soft” ionization techniques (i.e., ESI
and MALDI) [35,36], ion funnels were shown to collect, confine,
∗ Corresponding author. Tel.: +1 978 663 3600. and efficiently transfer ions through differentially pumped regions
E-mail address: mel.park@bruker.com (M.A. Park). of a mass spectrometer. Incorporating ion funnels at the front and

http://dx.doi.org/10.1016/j.ijms.2016.03.004
1387-3806/© 2016 Elsevier B.V. All rights reserved.
J.A. Silveira et al. / International Journal of Mass Spectrometry 413 (2017) 168–175 169

back of a drift tube has also improved the efficiency of ion injection vacuum stage of a prototype impact ESI-QqTOF (Bruker Daltonics,
as well as the collection of the radially diffuse ion swarm [37]. Billerica, MA) mass spectrometer. As depicted in Fig. 1a, the appa-
A subsequent ion funnel design featured a dedicated trapping ratus is comprised of a set of electrodes that form three regions:
region for enhanced utilization efficiency when used to pulse ions the entrance funnel, the TIMS tunnel, and the exit funnel. Ions and
into a drift tube [38]. Still, drift tube IMS demands that the initial N2 (g) are introduced to the first vacuum stage through the capil-
temporal width of the ion packet be very small with respect to the lary. During operation, ions are deflected into the entrance funnel
drift time (∼1%) to achieve high resolving power (R ≥ 50). For typical while gas may be pumped through a port positioned opposite to
drift times of ∼50 ms, this limits injection times to a few tenths of the capillary exit. Alternatively, during TIMS operation, gas flow
milliseconds and places a major constraint on the instrument duty through the port is partially or completely restricted such that
cycle. Though proof-of-principle temporal multiplexing strategies gas is diverted through the TIMS tunnel and is pumped through
have been shown to increase the ion throughput [39], challenges a secondary port located in the exit funnel region. Specific details
due to peak aliasing and the finite duty cycle limit of ∼50% remain. of the first-generation TIMS instrumentation can be found in the
In addition, temporal multiplexing strategies eliminate the ability literature [45,46].
to perform LC-IMS-MS and LC-IMS-MS/MS due to the need to recon- The second-generation TIMS funnel, featuring a 96 mm-long
struct the IMS-MS dataset over a complete multiplexing sequence. tunnel, was tested on a prototype microTOF (Bruker Daltonics, Bil-
A more recent report described the use of helium (as an alternative lerica, MA) mass spectrometer. In this design, the entrance and exit
to nitrogen) in the trapping region [40]. Owing to the greater mobil- funnel regions are identical to the original design; however, the
ity of ions in a lighter gas, an increase in the charge density of the tunnel length is increased by roughly a factor of two over that of
injected ion packet was observed, resulting in a sensitivity improve- the first-generation prototype (46 mm). Doubling the length of the
ment of more than an order of magnitude. Though more complete tunnel allows for the inclusion of a new “ion accumulation trap”
ejection of the accumulated ions was reported when using helium, upstream of the analysis region. The accumulation trap enables
the IMS duty cycle (1.6–33%) was still far from 100% since the capac- a mode of operation wherein incoming ions from the entrance
ity of the ion funnel/trap was not sufficient to store ions for the funnel can be accumulated, while simultaneously, a previously
entire duration of the 60 ms drift time. accumulated population of ions is mobility-analyzed in the “anal-
Introduced in 2011 by Park and coworkers, trapped ion mobility ysis region”. Inasmuch as ions are continuously accumulated, this
spectrometry (TIMS) was developed as a highly versatile alterna- mode of operation eliminates the need to block ions at the deflec-
tive to conventional drift tube IMS [41–43]. The general principles tion plate and, by definition, allows for a duty cycle of 100%.
of TIMS [44] and a theory [45,46] have been recently reported. In The second-generation TIMS funnel has three potential modes of
TIMS, ions are accumulated and stored according to their mobil- operation: (1) MS-only/transmission mode wherein the DC fields
ity in a manner analogous to the condensed phase electrophoretic and gas flow continuously guide the continuous ion beam to the
technique termed “electric field gradient focusing”. However, TIMS detector, (2) serial TIMS operation wherein ions are sequentially
also contains an elution step wherein ions are released and sepa- accumulated and analyzed, and (3) parallel accumulation TIMS
rated along an electric field gradient (EFG) plateau. The complete operation wherein ion accumulation and analysis occur in parallel.
analysis sequence occurs inside an ion funnel located in the first In MS-only mode, ions are simply transmitted from the entrance
vacuum stage of the instrument. Among other variables, the resolv- funnel to the exit funnel and subsequently, downstream to the
ing power has been shown to be dependent on the scan rate of the mass spectrometer without mobility analysis. In serial TIMS mode,
electric field (ˇ) during the elution step, as well as the drag force act- ions are accumulated and mobility-analyzed sequentially, as previ-
ing on the ions. In agreement with theory, experimental conditions ously described [45]. However, with the second-generation design,
that maximize the work done on ions—longer scans that maxi- one may take advantage of the additional storage capacity afforded
mize the “effective path length” and/or elevated pressures/higher by the longer tunnel [49]. Finally, in parallel accumulation mode,
gas velocities that increase the drag force—have yielded resolving one may achieve a 100% duty cycle, as outlined above. How-
power up to ∼230 for singly charged ions, ∼270 for doubly charged ever, by using the deflection plate to block ions for part of the
peptide ions [47] and ∼300 for multiply charged protein ions [48]. experimental cycle, it is also possible to employ longer analysis
Though exceptional performance has been demonstrated, thus times (while maintaining relatively short accumulation times to
far, TIMS has operated on the basis of an experimental sequence avoid charge capacity issues) when high resolution IMS analysis is
wherein ions are sequentially accumulated and analyzed. desired.
In the present work, we report a new spectrometer and The entrance and exit funnel regions of the second-generation
operational scheme that significantly improves the overall TIMS- spectrometer were pumped using a Roots-type multistage dry vac-
TOF MS instrument performance. The new design features a uum pump (∼600 m3 /h, Alcatel, 601B Roots Blower) backed by an
tunnel—roughly twice the length of the original design—that con- Ebara SA-20 mechanical pump that yielded a pressure of 3.1 and
tains an ion accumulation trap upstream of the analysis region. 0.9 mbar in each respective region. The pressure difference gener-
Spatially decoupling the accumulation from the analysis region ated across the tunnel has been shown to yield gas velocities on
allows the events to occur simultaneously, thus enabling TIMS the order of ∼150 m/s [46]. Ions exiting the TIMS funnel are guided
operation at 100% duty cycle. As demonstrated herein, operating into the TOF accelerator operated at ∼11 Hz. Both instruments were
the new spectrometer in this manner yields high TIMS resolving equipped with an electrospray ionization (ESI) source operated
power, high ion utilization efficiencies, and marked improvements with a 5 L/min N2 (g) flow rate. For all experiments, the shape of
in the TOF signal intensities, making this approach particularly well the EFG rising edge with respect to position followed a square root
suited for high throughput analyses and other applications requir- function that has been previously shown to yield equivalent axial
ing high sensitivity. storage space across the mobility range [49].
To characterize the instrument performance, a 1 pmol/␮L tryp-
tic digest of bovine serum albumin (BSA, Protea Biosciences,
2. Experimental Morgantown, WV) was directly infused at ∼150 ␮L/min from a
50:50 mixture of water:acetonitrile containing 0.1% formic acid.
A schematic representation of the first-generation spectrome- In addition, low concentration tuning mix (Agilent Technologies,
ter, its operation, and the coordinate system referred to herein is Santa Clara, CA) was used to determine the resolving power by Eq.
shown in Fig. 1a. The TIMS funnel was incorporated into the first (1), efficiency by Eq. (2), and signal enhancement factor by Eq. (3)
170 J.A. Silveira et al. / International Journal of Mass Spectrometry 413 (2017) 168–175

Fig. 1. Comparison of the original 46 mm-long tunnel and serial analysis scheme (a) with the 96 mm-long TIMS tunnel and parallel accumulation scheme (b). In (c), the duty
cycle is plotted as a function of accumulation time for serial (purple line) and parallel accumulation schemes (dotted blue line). In the serial mode of operation, the sum of
the elution and trap steps were assumed to be 50 ms. Note that the transfer step in (b) is very small (<1%) compared to the accumulation/elution step. Furthermore, ions
entering the accumulation trap during the transfer step are immediately transmitted to the analyzer, thus allowing for a duty cycle which is truly 100%.

for TIMS experiments where a broad mass range (m/z 622–2722) funnel/tunnel regions by pulsing the potential of the deflector lens
was analyzed: to an attractive potential. During this time, trapped ions acquire
an equilibrium position along the rising edge and are held for a
R = K/K (1)
user-defined time period, typically, on the order of a few milli-
SignalTIMS 1 seconds. During the elution step, the magnitude of the EFG profile
Efficiency = 100% · · (2) is decreased from an initial value at a fixed scan rate such that
SignalTrans Duty Cycle
ions of progressively higher K elute from the analyzer. Given that
ITIMS scanning the EFG at a slower rate is known to improve the resolv-
Enhancement factor = (3)
ITrans ing power, the ability to vary the time during the accumulation
In Eqs. (1)–(3), R is the resolving power and K is the ion mobility and scanning steps makes it possible to trade speed/duty cycle for
coefficient. ITIMS and ITrans are taken to be the maximum intensity resolving power.
observed for a species during TIMS and the constant intensity of
a species contained in the continuous beam in MS-only experi-
ments, respectively. It is important to recognize that efficiency, as 3.2. Theoretical charge capacity estimation
defined herein, is a measure of the complete system (particularly
with respect to tuning of the entrance funnel, TIMS tunnel, and exit It is, of course, natural that any device used to store ions should
funnel) when operating in parallel accumulation TIMS mode versus have a limited charge capacity. Therefore, part of our effort in
the system as a whole in MS-only mode. SignalTIMS , used for cal- understanding TIMS must be to determine an approximate charge
culating the efficiency, is taken to be the area under an IMS-MS capacity and the consequences of exceeding this finite limit. In
peak whereas SignalTrans is the chromatographic area observed in considering the operation of Paul traps, one may consider the stor-
MS-only mode during the time required for a TIMS experiment. age capacity (i.e., the maximum number of charges that may be
confined in the physical dimensions of the trap) and the analyt-
3. Results and discussion ical capacity (i.e., the maximum number of charges that may be
confined without substantially degrading analytical performance).
3.1. Ion storage in TIMS Furthermore, it is known that in Paul traps, the distribution of
charges with respect to m/z has an influence on relative perfor-
As depicted in Fig. 1, a feature of TIMS is that the trapping posi- mance as a function of mass. That is, if the ion density is relatively
tion for a particular ion is related to the ion mobility coefficient. The high in a narrow m/z range, the performance of the analyzer in
drag force exerted by the gas flow on an ion in the TIMS tunnel is a that particular m/z range will be degraded relative to the per-
simple function of the ion’s mobility, qvg /K. Also, the counteracting formance outside of that range [50]. Similar statements can be
force from the DC electric field is a function of the axial position made with respect to the TIMS device. In the present work, we
(proportional to qz1/2 ) as implied by the plot in Fig. 1. Accordingly, emphasize the analytically useful charge capacity of the TIMS ana-
ions of low K are dragged by the gas to an axial position further into lyzer.
the tunnel where the strength of the electric field on the rising edge If one considers a line of charge trapped in a TIMS analyzer, an
is greater. Conversely, ions of higher K are trapped at a position near associated electric field strength can be estimated based on funda-
the tunnel entrance where the strength of the electric field on the mental principles. Consider that the rising edge of the EFG profile
rising edge is relatively weaker. (where the ions are stored according to their mobility) is ∼23 mm
A serial analysis sequence (see Fig. 1a) is comprised of accu- long and the plateau region (over which the ions must “elute”) is
mulation, blocking, and elution steps. The second blocking step ∼18 mm in length [46]. As an approximation, we therefore assume
involves preventing additional ions from entering the entrance 106 charges, stored with uniform density in a line 23 mm in length,
J.A. Silveira et al. / International Journal of Mass Spectrometry 413 (2017) 168–175 171

in free space (i.e., neglecting electrodes). The electric field strength


at radial distance, r, on the bisector of the line of charge is given by:
q 1
Er =  (4)
2εo r l2 + 4r 2
where q is the total charge on the line and l is the length of the line
of charges [51]. From Eq. (4), the electric field strength 2 mm from
the axis in the TIMS analyzer holding a million charges would be
roughly 0.6 V/cm—roughly two orders of magnitude less than the
TIMS analytical field strength.
It is useful to point out that as the ions elute from the TIMS
analyzer, they must traverse the plateau region of the EFG profile.
As the ions cross the plateau, they are spatially separated from the
stored ions and the influence of the stored ions on the eluting ions
is therefore diminished. Assuming the same conditions as given
above, the electric field on the EFG plateau due to the charges stored
on the rising is given by:
q
1 1

Ez = − (5)
4εo l lo l + lo
where lo is the distance from the near end of the line of stored
charges to the position on the plateau. By Eq. (5), the electric
field strength at a position 6 mm onto the plateau resulting from
a million charges stored on the rising edge would be roughly
0.09 V/cm—again, roughly two orders of magnitude less than the
TIMS analytical field strength.
These two field strength estimates suggest that the storage of
106 to 107 charges should be possible without having a major
impact on the analytical performance of the TIMS analyzer. Fur-
thermore, the estimates suggest that the radial field strength
experienced during storage will have more of an influence on the
ions than the axial field strength experience during elution. Thus,
one might expect it to be more likely that ions are lost radially
during storage rather than being shifted in their observed elution
voltage.
Fig. 2. Representative 2D TIMS-MS plot of tuning mix (a) and a BSA digest (b)
3.3. Performance evaluation of the original TIMS analyzer acquired on the 46 mm-long analyzer.

The charge capacity and performance of the original 46 mm-long


becomes nonlinear indicating that the capacity at the correspond-
tunnel was evaluated using tuning mix and tryptic peptides from
ing axial trapping position has been reached. Fig. 3c shows that a
a BSA digest (see Fig. 2). Fig. 2a reveals separation of the homol-
similar phenomenon is also observed for m/z 288 (2+) and m/z 1258
ogously substituted fluorinated triazatriphosphorine (tuning mix)
(2+). However, m/z 711 (2+) and m/z 942 (2+)—which reside in a
ions from other singly charged background ions along distinctive
particularly dense region of the 2D TIMS-MS spectrum—decrease
mobility/mass trendlines. Similarly, Fig. 2b demonstrates trendline
in abundance with additional accumulation time as a result of com-
dispersion of peptide ions that differ in their charge state (i.e., sepa-
petition for space in the trap by neighboring ions. Generally, as ion
ration of singly from multiply charged ions). The capacity of the trap
traps reach their capacity, loss of low m/z ions has been attributed
was determined from a plot of the measured ion intensities versus
to compression of the stable m/z range. Alternatively, loss of high
the accumulation time. Fig. 3a shows that as expected, the inten-
m/z ions often results from radial stratification effects [38]. That is,
sities for tuning mix ions increase linearly until the trap reaches
when the radius of a particular ion exceeds the radius of the con-
capacity whereupon the accumulation rate decreases. For tuning
fining potentials produced by the trap geometry, the ion will no
mix ions, the data indicate that this point occurs at ∼150 ms. Given
longer be stable and will be neutralized upon striking the surface of
that the sample produced 7 pA of continuous ion current (measured
an electrode [38,52]. In a TIMS experiment, ions of near-identical
downstream of the TIMS analyzer) and assuming a trapping effi-
mobility but dissimilar m/z are trapped in the same axial space;
ciency of 70% (see Section 3.2), the number of singly charged ions
thus, overfilling the trap can result in preferential loss of high m/z
stored is estimated to be ∼4.6 × 106 —in agreement with the first
ions that occupy the largest radii.
principles estimation of the total charge capacity discussed above.
A similar evaluation is more complex when tracking individual
ion intensities found in a tryptic BSA digest sample. At 1 pmol/␮L, 3.4. Performance evaluation of the parallel accumulation
the sample was found to produce a beam current roughly ∼20 pA measurement scheme
measured downstream of the TIMS analyzer. The relatively higher
beam current (roughly threefold greater compared to tuning mix) From the standpoint of maintaining a high duty cycle in a
and the presence of multiply charged ions together can, under serial analysis scheme, it is advantageous for the accumulation
some circumstances, produce effects that arise from mobility- time to be large with respect to the scan time. However, TIMS
based storage. Fig. 3b shows that singly charged ion intensities theory and experimental results have established that reduc-
largely increase upon extending the accumulation time up to ing the rate of the EFG scan improves
 the resolving power (i.e.,
200 ms, though the accumulation rate for m/z 791 (1+), for example, resolving power is proportional to 4 1/ˇ) [45,46]. Thus, reducing
172 J.A. Silveira et al. / International Journal of Mass Spectrometry 413 (2017) 168–175

approach was previously successfully implemented, serial analysis


still suffers from the same principal duty cycle limitation outlined
above. Thus, as an alternative, the tunnel was segmented into two
regions configured with independent voltage control in a new mode
of operation.
As depicted in Fig. 1b, the new parallel accumulation mode
allows for the accumulation and analysis steps to occur simulta-
neously. Ions produced in the source are accumulated during the
analysis of a previously accumulated population, thus eliminating
the need to block ions using the deflector. Most notably, if the scan
time is equivalent to the accumulation time, the duty cycle can be
100% (see Fig. 1c). Note that in a serial analysis mode, the duty
cycle for equivalent operational parameters would only yield 50%.
Moreover, the new approach maintains the inherent flexibility to
trade off speed/duty cycle and/or mobility range when additional
resolving power is required.
A number of experiments were performed to explore the per-
formance of the new spectrometer. Fig. 4a contains a plot of the
resolving power as a function of the accumulation/analysis time
when operating TIMS at 100% duty cycle. The data indicate that as
expected from TIMS theory, the resolving power initially increases
upon extending the analysis time following a fourth-root depend-
ence. However, at analysis times exceeding 30 ms, the gain in
resolving power typically observed as a result of extending the scan
time appears to be largely offset by space charge effects. Though
most of the separation is achieved in a 30 ms scan, Fig. 4b shows that
the trap still has the capacity to store additional ions, evidenced by
the linear increase in the plot of the average peak area as a function
of accumulation time.
Next, to determine the extent of axial space charge broadening
that results from increasing the number of ions stored, the analysis
time was fixed at 120 ms while the duty cycle was varied by block-
ing ions entering the accumulation trap for a specified fraction of
the analysis sequence. In doing so, the average ion intensity across
the mass range was found to scale linearly with the duty cycle up to
100% (see Fig. 4c). At duty cycles between 1% and 10%, the resolv-
ing power was found to be fairly constant indicating that when the
number of ions stored is relatively low, broadening of the ion swarm
due to space charge effects is negligible (see Fig. 4d). However, upon
increasing the duty cycle (and thus the number of ions stored)
Fig. 3. Tuning mix (a) and select singly (b) and doubly (c) charged ions from a beyond 10%, a decrease in the resolving power (up to 30%) was
BSA digest plotted as a function of the accumulation time. Note that the data were observed when the spectrometer was operated at 100% duty cycle.
acquired on the 46 mm-long analyzer. Nevertheless, when operating at 100% duty cycle, the average TIMS
resolving power across the mobility range (R = 62) is still equiva-
lent to high resolution drift tube IMS separations [54]. Accordingly,
ˇ allows for improved separation performance, but does so at the a representative two-dimensional TIMS-MS plot is shown in Fig. 4e
cost of a reduced duty cycle and/or mobility range. Past experi- demonstrating separation comparable to serial mode of operation
mental data demonstrate that employing a scan time of 50 ms, it (Fig. 2a).
is feasible to achieve a resolving power exceeding 50 for a wide It is important to note that 100% duty cycle was demonstrated
range of singly charged ions [45]; assuming an accumulation time at an accumulation time of 120 ms and that because the accumula-
of 150 ms, a duty cycle of ∼70% can be achieved. However, as shown tion/analysis time is relatively long, charge capacity, for analytical
in Fig. 1c, a duty cycle of 100% is only asymptotically approached purposes, has been exceeded. In alternative timing methods, it is
as the accumulation time is extended further. Also, as discussed in possible to achieve a 100% duty cycle without effects due to charge
Section 3.1, the accumulation time has a practical limit set by the capacity. For example, limiting the accumulation/analysis time to
capacity of the trap and the ion flux produced by the source/sample. 30 ms would achieve 100% duty cycle and yield similar resolving
Thus, in practice, a serial measurement scheme cannot simul- power while avoiding charge capacity effects. Furthermore, higher
taneously yield a duty cycle of 100% and high TIMS resolving mobility resolution in combination with 100% duty cycle at a 30 ms
power. accumulation/analysis time can be achieved by limiting the mobil-
To overcome this principal limitation, we redesigned the tun- ity range analyzed. Limiting the mobility range results in a lower
nel region to incorporate an ion accumulation trap upstream of the scan rate, ˇ, which, as discussed previously, results in higher reso-
TIMS analyzer. Note that this second-generation spectrometer can lution.
still be operated in the serial mode if so desired. In serial mode, one Finally, to assess the efficiency of storing and transferring ions
may consider using the added length of tunnel for improved capac- during TIMS, the relative signals produced in continuous and
ity. That is, by simply extending the axial length of the rising edge trapping modes of operation were compared. Fig. 5a shows the
of the EFG profile, the charge capacity can be increased by a fac- efficiency at 100% duty cycle for various accumulation/analysis
tor of three over the first-generation prototypes [53]. Though this times plotted as a function of m/z. The data reveal no statistically
J.A. Silveira et al. / International Journal of Mass Spectrometry 413 (2017) 168–175 173

Fig. 4. (a) Average resolving power for tuning mix ions (m/z 622–2722) versus accumulation/analysis time for operation of the 96 mm-long tunnel at 100% duty cycle. The
dotted line represents an extrapolation of the separation performance expected based on fourth-root dependence of the data acquired at relatively short accumulation/analysis
times (≤13 ms). The average TIMS peak area versus accumulation/analysis time is shown in (b). In (c) and (d), the duty cycle was varied using a fixed analysis time of 120 ms to
study the signal intensity (c) and the resolving power (d). Panel (e) contains a 2D TIMS-MS plot acquired on the 96 mm-long tunnel using the parallel accumulation/analysis
scheme operating at 100% duty cycle (120 ms accumulation/analysis time).

significant dependence when the experiment was performed at varied timescales indicated that (1) at 85 ms, the total charge capac-
accumulation/analysis times between 20 and 85 ms. Rather, the ity of the trap has not yet been reached and (2) that m/z-dependent
average trapping efficiency was found to be ∼70% for all timescales stability in the trap is not the primary factor that contributes to
investigated. That the efficiency remains constant for each m/z at ion losses. Rather, the m/z dependence observed in Fig. 5a is likely
174 J.A. Silveira et al. / International Journal of Mass Spectrometry 413 (2017) 168–175

in parallel accumulation mode in an identical 47 ms time period.


It is clear that all the ions that would have otherwise appeared
as a continuous, low level signal now appear in a relatively high
intensity peak for a short duration. When operating at 100% duty
cycle, the apex of the mobility peak has a signal intensity 80 times
greater than that observed in transmission mode. The ratio of these
signals, termed the enhancement factor (see Fig. 5b), was found to
be between a factor of 60 and 120.
It is worth mentioning that it is because the ions are retained
and transmitted with relatively high efficiency, because a high
duty cycle is achieved, and because the IMS peaks are temporally
narrow, the signal intensities during ion elution are substantially
higher than in transmission mode. Importantly, assuming a con-
stant level of noise, this implies that the signal-to-noise ratio is
also concomitantly increased by the same factor. Furthermore, as
TIMS yields a mobility-based separation, we can also expect a
reduction in chemical background, further improving the signal-to-
noise ratio relative to the MS-only operational mode. Also, because
ions are eluted over time, a duty cycle enhancement in down-
stream devices can also be achieved, as recently described by Mann
et al. in a method termed “Parallel Accumulation SErial Fragmen-
tation” (PASEF) [31]. While space charge effects observed herein
were found to limit further gains in the resolving power for scan
times exceeding 30 ms, the capacity of the trap was found to effi-
ciently store additional ions without loss in performance. This
ability should prove useful in PASEF methods such that additional
precursor ions can be selected from a complex mixture in a single
experiment.

4. Conclusions

A new trapped ion mobility spectrometry (TIMS) apparatus has


been developed and evaluated. In addition to a TIMS analyzer, the
new spectrometer includes an upstream ion accumulation trap. The
new trap can be used to accumulate a group of ions while a pre-
viously accumulated group of ions is simultaneously analyzed. By
accumulating and analyzing ions in parallel rather than in series, a
duty cycle of 100% was demonstrated. The spectrometer was shown
to simultaneously trap a broad range of ions (m/z 622–2722) with
an average efficiency of ∼70%. Trapping efficiencies were found to
be constant over a range of accumulation/analysis times from 20 up
to 85 ms where an average resolving power of ∼60 was achieved. It
was found that, as currently constructed, the TIMS trap has an ana-
lytically useful charge capacity of ∼5 × 106 charges. The capacity
Fig. 5. TIMS efficiency (a) and enhancement factor (b) at 100% duty cycle for various
accumulation/analysis times between 20 and 85 ms plotted as a function of m/z. In and ion beam current together set a limit on the accumulation time.
(c), the signal observed for m/z 1222 in parallel accumulation mode is normalized Compared to MS-only operation, the combination of high mobil-
to the signal observed in MS-only (transmission) mode. ity resolving power, high ion utilization efficiency, and relatively
fast scan times achieved with TIMS is shown to improve the mass
spectral peak intensities by an average factor of ∼80.
the result of differences in ion transmission injecting ions into and
capturing ions out of the TIMS tunnel.
In contrast to MS-only operation, ion packets are temporally dis- References
persed based upon their mobility during TIMS. As a consequence,
[1] W.S. Barnes, D.W. Martin, E.W. McDaniel, Mass spectrographic identification of
operating at very short accumulation/analysis times (<20 ms) can
the ion observed in hydrogen mobility experiments, Phys. Rev. Lett. 6 (1961)
produce ion packets with narrow temporal peak widths that may 110–111.
not overlap with the TOF acceleration pulse. Alternatively, at accu- [2] C.H. Bloomfield, J.B. Hasted, New technique for the study of ion-atom inter-
mulation/analysis times >20 ms, the average peak widths were change, Discuss. Farad. Soc. 37 (1964) 176–184.
[3] H.E. Revercomb, E.A. Mason, Theory of plasma chromatography/gaseous elec-
determined to be ∼400 ␮s such that more than four TOF spectra trophoresis. Review, Anal. Chem. 47 (7) (1975) 970–983.
were collected across the width of the TIMS peak. For accumu- [4] W.E. Steiner, C.S. Harden, F. Hong, S.J. Klopsch, H.H. Hill Jr., V.M. McHugh,
lation/analysis times between 20 and 85 ms, the ion intensities Detection of aqueous phase chemical warfare agent degradation products by
negative mode ion mobility time-of-flight mass spectrometry [IM(tof)MS], J.
measured at the centroid of peak were found to greatly exceed the Am. Soc. Mass Spectrom. 17 (2) (2006) 241–245.
intensity for the same m/z ratio in the continuous beam. This effect [5] A. McKenzie-Coe, J.D. DeBord, M. Ridgeway, M. Park, G. Eiceman, F.
is demonstrated clearly in Fig. 5c wherein the signal for m/z 1222 is Fernandez-Lima, Lifetimes and stabilities of familiar explosive molecular
adduct complexes during ion mobility measurements, Analyst 140 (16) (2015)
measured with and without TIMS analysis. The transmission mode 5692–5699.
signal is shown over a period of ∼47 ms and is normalized to an [6] J. Gidden, M. Bowers, A. Jackson, J. Scrivens, Gas-phase conformations of cation-
intensity of one. The additional traces show the peak for m/z 1222 ized poly(styrene) oligomers, J. Am. Soc. Mass Spectrom. 13 (5) (2002) 499–505.
J.A. Silveira et al. / International Journal of Mass Spectrometry 413 (2017) 168–175 175

[7] S. Trimpin, D.E. Clemmer, Ion mobility spectrometry/mass spectrometry snap- [31] F. Meier, S. Beck, N. Grassl, M. Lubeck, M.A. Park, O. Raether, M. Mann, Parallel
shots for assessing the molecular compositions of complex polymeric systems, accumulation-serial fragmentation (PASEF): multiplying sequencing speed and
Anal. Chem. 80 (23) (2008) 9073–9083. sensitivity by synchronized scans in a trapped ion mobility device, J. Proteome
[8] F.A. Fernandez-Lima, C. Becker, K. Gillig, W.K. Russell, M.A.C. Nascimento, D.H. Res. 14 (12) (2015) 5378–5387.
Russell, Experimental and theoretical studies of (CsI)n Cs+ cluster ions pro- [32] P. Benigni, C.J. Thompson, M.E. Ridgeway, M.A. Park, F. Fernandez-Lima, Tar-
duced by 355 nm laser desorption ionization, J. Phys. Chem. A 112 (44) (2008) geted high-resolution ion mobility separation coupled to ultrahigh-resolution
11061–11066. mass spectrometry of endocrine disruptors in complex mixtures, Anal. Chem.
[9] C. Bleiholder, N.F. Dupuis, T. Wyttenbach, M.T. Bowers, Ion mobility-mass 87 (8) (2015) 4321–4325.
spectrometry reveals a conformational conversion from random assembly to [33] S.A. Shaffer, K. Tang, G.A. Anderson, D.C. Prior, H.R. Udseth, R.D. Smith, A
␤-sheet in amyloid fibril formation, Nat. Chem. 3 (2) (2011) 172–177. novel ion funnel for focusing ions at elevated pressure using electrospray ion-
[10] J.A. Silveira, K.A. Servage, C.M. Gamage, D.H. Russell, Cryogenic ion mobility- ization mass spectrometry, Rapid Commun. Mass Spectrom. 11 (16) (1997)
mass spectrometry captures hydrated ions produced during electrospray 1813–1817.
ionization, J. Phys. Chem. A 117 (5) (2013) 953–961. [34] S.A. Shaffer, D.C. Prior, G.A. Anderson, H.R. Udseth, R.D. Smith, An ion funnel
[11] K.A. Servage, J.A. Silveira, K.L. Fort, D.H. Russell, Evolution of hydrogen bond interface for improved ion focusing and sensitivity using electrospray ioniza-
networks in protonated water clusters H+ (H2 O)n (n = 1 to 120) studied by tion mass spectrometry, Anal. Chem. 70 (19) (1998) 4111–4119.
cryogenic ion mobility-mass spectrometry, J. Phys. Chem. Lett. 11 (2014) [35] J.B. Fenn, M. Mann, C.K. Meng, S.F. Wong, C.M. Whitehouse, Electrospray ion-
1825–1830. ization for mass spectrometry of large biomolecules, Science 246 (1989) 64–71.
[12] K.A. Servage, K.L. Fort, J.A. Silveira, L. Shi, D.E. Clemmer, D.H. Russell, Unfolding [36] M. Karas, F. Hillenkamp, Laser desorption ionization of proteins with molecular
of hydrated alkyl diammonium cations revealed by cryogenic ion mobility- masses exceeding 10,000 Daltons, Anal. Chem. 60 (20) (1988) 2299–2301.
mass spectrometry, J. Am. Chem. Soc. 137 (28) (2015) 8916–8919. [37] K. Tang, A.A. Shvartsburg, H.N. Lee, D.C. Prior, M.A. Buschbach, F. Li, A.V.
[13] T. Wyttenbach, G. von Helden, M.T. Bowers, Gas-phase conformation of biolog- Tolmachev, G.A. Anderson, R.D. Smith, High-sensitivity ion mobility spectrome-
ical molecules: Bradykinin, J. Am. Chem. Soc. 118 (35) (1996) 8355–8364. try/mass spectrometry using electrodynamic ion funnel interfaces, Anal. Chem.
[14] B.T. Ruotolo, J.L.P. Benesch, A.M. Sandercock, S.J. Hyung, C.V. Robinson, Ion 77 (10) (2005) 3330–3339.
mobility-mass spectrometry analysis of large protein complexes, Nat. Protoc. [38] B.H. Clowers, Y.M. Ibrahim, D.C. Prior, W.F. Danielson, M.E. Belov, R.D. Smith,
3 (7) (2008) 1139–1152. Enhanced ion utilization efficiency using an electrodynamic ion funnel trap
[15] M.F. Bush, Z. Hall, K. Giles, J. Hoyes, C.V. Robinson, B.T. Ruotolo, Collision cross as an injection mechanism for ion mobility spectrometry, Anal. Chem. 80 (3)
sections of and their complexes: a calibration framework and database for gas- (2008) 612–623.
phase structural biology, Anal. Chem. 82 (22) (2010) 9557–9565. [39] M.E. Belov, M.A. Buschbach, D.C. Prior, K. Tang, R.D. Smith, Multiplexed ion
[16] N.A. Pierson, L. Chen, S.J. Valentine, D.H. Russell, D.E. Clemmer, Number of mobility spectrometry-orthogonal time-of-flight mass spectrometry, Anal.
solution states of Bradykinin from ion mobility and mass spectrometry mea- Chem. 79 (6) (2007) 2451–2462.
surements, J. Am. Chem. Soc. 133 (35) (2011) 13810–13813. [40] Y.M. Ibrahim, S.V.B. Garimella, A.V. Tolmachev, E.S. Baker, R.D. Smith, Improving
[17] T. Wyttenbach, M.T. Bowers, Structural stability from solution to the gas phase: ion mobility measurement sensitivity by utilizing helium in an ion funnel trap,
native solution structure of ubiquitin survives analysis in a solvent-free ion Anal. Chem. 86 (11) (2014) 5295–5299.
mobility-mass spectrometry environment, J. Phys. Chem. B 115 (42) (2011) [41] M.A. Park, Apparatus and method for parallel flow ion mobility spectrometry
12266–12275. combined with mass spectrometry, US Patent 8,288,717 (2010).
[18] Z. Hall, A. Politis, M.F. Bush, L.J. Smith, C.V. Robinson, Charge-state dependent [42] F. Fernandez-Lima, D.A. Kaplan, M.A. Park, Integration of trapped ion mobil-
compaction and dissociation of protein complexes: insights from ion mobility ity spectrometry with mass spectrometry, Rev. Sci. Instrum. 82 (2011)
and molecular dynamics, J. Am. Chem. Soc. 134 (7) (2012) 3429–3438. 126106.
[19] E.R. Schenk, M.E. Ridgeway, M.A. Park, F. Leng, F. Fernandez-Lima, Isomerization [43] F. Fernandez-Lima, D. Kaplan, J. Suetering, M. Park, Gas-phase separation using
kinetics of AT hook decapeptide solution structures, Anal. Chem. 86 (2) (2013) a trapped ion mobility spectrometer, Int. J. Ion Mobil. Spectrom. 14 (2–3) (2011)
1210–1214. 93–98.
[20] L. Chen, S.H. Chen, D.H. Russell, An experimental study of the solvent- [44] D.R. Hernandez, J.D. DeBord, M.E. Ridgeway, D.A. Kaplan, M.A. Park, F.
dependent self-assembly/disassembly and conformer preferences of grami- Fernandez-Lima, Ion dynamics in a trapped ion mobility spectrometer, Analyst
cidin A, Anal. Chem. 85 (16) (2013) 7826–7833. 139 (2014) 1913–1921.
[21] S. Warnke, C. Baldauf, M.T. Bowers, K. Pagel, G. von Helden, Photodissociation of [45] K. Michelmann, J. Silveira, M. Ridgeway, M. Park, Fundamentals of trapped ion
conformer-selected ubiquitin ions reveals site-specific cis/trans isomerization mobility spectrometry, J. Am. Soc. Mass Spectrom. 26 (1) (2015) 14–24.
of proline peptide bonds, J. Am. Chem. Soc. 136 (29) (2014) 10308–10314. [46] J.A. Silveira, K. Michelmann, M.E. Ridgeway, M. Park, Fundamentals of trapped
[22] K.J. Pacholarz, P.E. Barran, Distinguishing loss of structure from subunit dis- ion mobility spectrometry. Part II. Fluid dynamics, J. Am. Soc. Mass Spectrom.
sociation for protein complexes with variable temperature ion mobility mass (2016), http://dx.doi.org/10.1007/s13361-015-1310-z (in press).
spectrometry, Anal. Chem. 87 (12) (2015) 6271–6279. [47] J.A. Silveira, M.E. Ridgeway, M.A. Park, High resolution trapped ion mobility
[23] J.C. May, J.A. McLean, A uniform field ion mobility study of melittin and impli- spectrometery of peptides, Anal. Chem. 86 (12) (2014) 5624–5627.
cations of low-field mobility for resolving fine cross-sectional detail in peptide [48] M.E. Ridgeway, J.A. Silveira, J.E. Meier, M.A. Park, Microheterogeneity within
and protein experiments, Proteomics 15 (16) (2015) 2862–2871. conformational states of ubiquitin revealed by high resolution trapped ion
[24] W. Cui, H. Zhang, R.E. Blankenship, M.L. Gross, Electron-capture dissociation mobility spectrometry, Analyst 140 (20) (2015) 6964–6972.
and ion mobility mass spectrometry for characterization of the hemoglobin [49] M.A. Park, N. Park, J.A. Meier, M.E. Ridgeway, J.A. Silveira, Equal opportunity ion
protein assembly, Protein Sci. 24 (8) (2015) 1325–1332. storage: electric field optimization for complex mixture analysis by trapped ion
[25] S.J. Allen, K. Giles, T. Gilbert, M.F. Bush, Ion mobility mass spectrometry of pep- mobility spectrometry, in: Proceedings from the 63rd Annual American Society
tide, protein, and protein complex ions using a radio-frequency confining drift for Mass Spectrometry Conference, St. Louis, MO, May 31–June 4, 2015.
cell, Analyst 141 (3) (2016) 884–891. [50] D.A. Kaplan, R. Hartmer, A. Brekenfeld, J. Franzen, M. Schubert, An examination
[26] S. Tenzer, A. Moro, J. Kuharev, A.C. Francis, L. Vidalino, A. Provenzani, P. Macchi, of the physics of the high capacity trap, in: R.E. March, J.F.J. Todd (Eds.), Practical
Proteome-wide characterization of the RNA-binding protein RALY-interactome Aspects of Trapped Ion Mass Spectrometry, CRC Press, Boca Raton, FL, 2010, pp.
using the in vivo-biotinylation-pulldown-quant (iBioPQ) approach, J. Proteome 593–617.
Res. 12 (6) (2013) 2869–2884. [51] D. Halliday, R. Resnick, Physics: Part 2, John Wiley and Sons, New York, NY,
[27] V. Parviainen, S. Joenvaara, N. Tohmola, R. Renkonen, Label-free mass spec- 1960.
trometry proteome quantification of human embryonic kidney cells following [52] A.V. Tolmachev, H.R. Udseth, R.D. Smith, Modeling the ion density distribution
24 hours of sialic acid overproduction, Proteome Sci. 11 (1) (2013) 11–38. in collisional cooling RF multipole ion guides, Int. J. Mass Spectrom. 222 (2003)
[28] R.P. Dator, K.W. Gaston, P.A. Limbach, Multiple enzymatic digestions and ion 155–174.
mobility separation improve quantification of bacterial ribosomal proteins [53] K. Michelmann, J. Bossmeyer, M.E. Ridgeway, J.A. Silveira, M. Mann, M.A. Park,
by data independent acquisition liquid chromatography–mass spectrometry, In improved capacity trapped ion mobility spectrometry, in: Proceedings from
Anal. Chem. 86 (9) (2014) 4264–4270. the 63rd Annual American Society for Mass Spectrometry Conference, St. Louis,
[29] S. Helm, D. Dobritzsch, A. Rodiger, B. Agne, S. Baginsky, Protein identifica- MO, May 31–June 4, 2015.
tion and quantification by data-independent acquisition and multi-parallel [54] J.C. May, C.R. Goodwin, N.M. Lareau, K.L. Leaptrot, C.B. Morris, R.T. Kurulugama,
collision-induced dissociation mass spectrometry (MSE ) in the chloroplast A. Mordehai, C. Klein, W. Barry, E. Darland, G. Overney, K. Imatani, G.C. Stafford,
stroma proteome, J. Proteomics 98 (2014) 79–89. J.C. Fjeldsted, J.A. McLean, Conformational ordering of biomolecules in the gas
[30] D. Helm, J.P.C. Vissers, C.J. Hughes, H. Hahne, B. Ruprecht, F. Pachl, A. Grzyb, K. phase: nitrogen collision cross sections measured on a prototype high reso-
Richardson, J. Wildgoose, S.K. Maier, H. Marx, M. Wilhelm, I. Becher, S. Lemeer, lution drift tube ion mobility-mass spectrometer, Anal. Chem. 86 (4) (2014)
M. Bantscheff, J.I. Langridge, B. Kuster, Ion mobility tandem mass spectrome- 2107–2116.
try enhances performance of bottom-up proteomics, Mol. Cell. Proteomics 13
(2014) 3709–3715.

You might also like